Generating a large number of customized and personalized outbound voice messages is a tedious task that requires substantial amount of manual processes including creating, communicating, and bookkeeping of a large number of sound files. It is also difficult to index such large number of sound files in a manner so that they can be used by a telephony system. Furthermore, using such a large number of sound files also makes it difficult to generate real-time outbound messages without considerable lag time between initiating a telephone connection with a target and delivering the sound file to the target. Thus, there is a need for a method and system that generates and deploys a large number of sound files in an efficient and effective manner.